Effects of formant enhancement on sentence recognition in noise: Influence of enhancement strategy, facemask muffling, and signal-to-noise ratio

Speech recognition in noise is particularly challenging when speech signals are muffled by face masks or presented at low signal-to-noise ratios (SNRs). Formant enhancement, a technique that selectively amplifies critical speech formants to improve intelligibility without excessive loudness, has shown promise in addressing these difficulties. The present study investigated how formant enhancement would affect sentence recognition in noise, focusing specifically on the enhancement strategy (second formant [F2] alone versus second and third formants [F2 & F3]), speech clarity (clear versus facemask muffled speech), and SNR (moderate and challenging SNRs). Twenty-one native English listeners with normal hearing completed sentence recognition tasks across three conditions (unmodified, F2-enhanced, and F2&F3-enhanced speech) in four-talker babble noise at moderate and challenging SNRs (0 & −3 dB). Results indicated that both enhancement strategies significantly improved sentence recognition performance relative to unmodified speech. However, the combined F2 & F3 enhancement consistently provided greater improvements than F2 enhancement alone, particularly with face-mask muffled speech and at challenging SNR conditions. These findings demonstrated the efficacy of formant enhancement, especially combined F2 & F3 amplification, as a potential enhancement approach for improving speech intelligibility under adverse listening conditions such as facemask use and challenging background noise.